Protocols.io – Life Sciences Protocol Repository | Hacker News Reader

Protocols.io – Life Sciences Protocol Repository

kickstarter.com

Years ago, when I hadn't yet completed the switch from the lab to computers, I used http://www.protocol-online.org/ from time to time, which is, in my understanding, a similar project. Though checking it out just now indicates that not much has been going on there for the last couple of years. irollboozers already mentioned http://openwetware.org/wiki/Main_Page as an alternative project with a wiki-style interface. Those sites attracted some participants but it seems they never really went mainstream - should ask the wet lab people about this to confirm.

As I'm working in Next Generation Sequencing (NGS) now after a stint in the software industry, I'm actually shocked by the lack of standardization and documentation in the wet lab, which makes a thorough statistical data analysis so much harder. Few use electronic lab books and even fewer associate those with experimental data.

I think that is what is really missing, a push to easy-to-use electronic lab books that use open, machine- and human-readable formats and a straightforward way of associating these with the generated data. Upload these documented and annotated raw data to a central repository and you create a computational biologist's wet dream.

But that would require a huge amount of work to create the necessary ontologies...

>I'm actually shocked by the lack of standardization and documentation in the wet lab

There's actually quite a lot of both of these. Here's the thing, we all have protocols specific to our labs or organisms. For many of us, our protocols are based on a canonical source (Which are now usually found in Nature's Prootcols) but optimized around what we do. Working with organics is rarely as easy as it might seem (especially if you're coming from the software side of things, I somwhat did too). I've digitized most of our protocols (some dating back to the 1950's) but for the most part we all refer to hardcopies of them. Electronic protocols aren't nearly as useful as print ones, especially since they're constantly tweaked. And most of us in academia freely share our optimized protocols with eachother -- you only have to ask.

There are a couple of Lab-notebook-in-the-cloud options but I've never really liked any of them. Instead I keep everything in Dropbox (for sharing with my colleagues) with versioning handled in a private git repo. I can commit experiment notes and checkin images. All of my notes are in markdown.

There are many ex-companies that tried to push for an electronic lab notebook (ELN). There are several other ELN companies today that seem to have gotten nowhere. That's due to a panel of reasons, most important of which - they are competing with a paper notebook. May not be digital, but getting information into the paper one is trivial. Getting it into all the others a huge barrier.

Some free protocols, but most are not/are available on a subscription basis, but http://www.springerprotocols.com/ could be a much suitable alternative. However, I'm sceptical that your company hasn't heard of them, perhaps it's an issue of available funds for it unfortunately.

There's a newsgroup (and gatewayed mailing list), running for god knows how long but at least since 1993 [1], called Methods and Reagents:

https://groups.google.com/forum/#!forum/bionet.molbio.methds...

Which has been used to share protocols, just like this, since the dawn of the web. It's a newsgroup, so hardly anyone uses it any more, of course. I guess it doesn't have a flat UI and an iPhone app.

At some point, some of the users of the group set up a wiki to capture protocols in a more permanent form:

http://methodsandreagents.pbworks.com/w/page/20806945/FrontP...

But it never took off like we hoped.

As irollboozers notes in another comment, there is nothing new under the sun. The selection criterion for any such effort is simply getting a critical mass of people involved and contributing protocols.

[1] http://schneider.ncifcrf.gov/methods/TIBS/nov93.txt

Not sure about if that will work.

Most of the time you can't change a thing in protocols if you want to have a minimal level of acceptance.

You would only need such a thing if you are developing a new protocol for something, which might happen a lot but I am not sure if it would be nice to be sharing not ready protocols on the web. Even if it would, people would still need to agree on a single one or something. Otherwise you will end up with 200000x DNA extraction protocols with their tiny customizations(Every lab technician has it's own way of doing every single step in some protocols...)

It would probably be nice as a in house access only system for universities and research facilities thought.

There's already Springer Protocols (http://www.springerprotocols.com/), not free, but IMHO, universities and research facilities already have it better than whatever free alternative may come up in the future. This is not the sort of project that requires only upfront costs, this is why I unfortunately don't think it will live for long, if it ever gets completely funded.

The changes, via a good user interface, will be visualized on step level. Very true that the UI is key here. But trust us - there will not be thousands of variants of a single protocol. There will be thousands of annotations, easily found for your species, on the single original published protocol.

The crowdsourced part is where this could really have a big impact, otherwise the technology is pretty standard. Crowdsourcing legitimate and replicable protocols is something others have tried (OpenWetWare, etc) but not quite scaled to Github like levels. Still, this seems to run into the same challenges that others before it faced.

I'll gladly pitch in some funds!

Indeed, the effort to create a central protocol repository is at least 20 years old. No one succeeded. The key part here is the crowdsourcing. Won't happen without a crowd. And won't happen unless a busy geneticist can share with a tap of a button. That's why for two years we have been focused on giving amazing tools to researchers - build the user base. We did not even begin building protocols.io and the checklist app right away because we know how hard it is to get traction. Also, we may just be doing this at the right time - on mobile. The protocol checklist app is the key ingredient that will make this happen.

I work in a lab that does high throughput screening, and I do develop internal web-based toolsets. After several years of iterations I can say that so far electronic lab notebooks never worked; there are even several commercial software available.

There are several reasons behind it:

1) Protocols are not a fixed predefined setup. One of the most important part of a protocol is the annotations on top. What went wrong, what did you change, etc. Today I have a repository of hand-written protocols (see [2]) that people print before going to the lab, write a date on it, and write comments on it. I cannot imagine this process of taking notes will be disrupted in near future, unless a very easy to use note taking system is there. IPad, LiveScribe (even wrote an app for it) et al are not there.

2) Writing protocols for lab people is cumbersome. Especially in our lab that has a protocol for every little process, to ensure reproducibility and standards. Asking them to sit behind computers (or mobile phone) to write is madness. The best that worked so far is I ask them to write and draw the process. Scan it, store in db, and print it with a click before experiment. They are happier than ever. Once done, I scan them back into db with date attached, and use evernote to search inside in case. The drawback is th editing part, but surprisingly they prefer to rewrite the whole thing rather than edit it behind the computer.

3) Protocols are subject to change, unless they are standard protocols (probably the market that this kickstarter is aiming). How should people store the revisions of the protocols? which one was used in the last experiment? did this modification work? think about a "git" system for protocols!

4) Lab people hate computers. They hate to use mobile phones to do anything than checking mails and calls. They love paper. Anything touchable. Taking mobiles into the lab is messy (google glass?). I have a feeling this can work with a geeky lab person but majority will oppose strongly.

5) Try to enter a protocol inside the tool. It will take ages until I have a smart protocol that may use non-standard agents inside. They need to work on a smarter way of entering protocols. Forms are too rigid.

6) I did an A/B test for some time, and what I found out was that pictures and text combination work great together, and minimized the common errors a lot. Printing a protocol, taking with them and ticking steps added another level of quality. So, adding hand-drawn / google images to the protocols may help besides an option for printing the protocol in the mobile app.

7) I love the common repository idea. The share button will be easy enough for even the laziest laborant.

I believe this is a good step toward the mess that is happening in the labs. Still the solution above currently is a local one to me; good enough for iteration 1 and will donate to push the idea forward.

I just wrote a long answer to "Will scientists use mobile in the lab?" http://blog.pubchase.com/was-henry-ford-an-idiot-will-scient...

Reply on news.ycombinator.com